1: \documentclass[9pt]{jdg-p-1-2}%
2: \usepackage{amsmath}%
3: \setcounter{MaxMatrixCols}{30}%
4: \usepackage{amsfonts}%
5: \usepackage{amssymb}%
6: \usepackage{graphicx}
7: %TCIDATA{OutputFilter=latex2.dll}
8: %TCIDATA{Version=5.50.0.2953}
9: %TCIDATA{LastRevised=Friday, June 02, 2006 22:51:16}
10: %TCIDATA{<META NAME="GraphicsSave" CONTENT="32">}
11: %TCIDATA{<META NAME="SaveForMode" CONTENT="1">}
12: %TCIDATA{BibliographyScheme=Manual}
13: %BeginMSIPreambleData
14: \providecommand{\U}[1]{\protect\rule{.1in}{.1in}}
15: %EndMSIPreambleData
16: \newtheorem{theorem}{Theorem}[section]
17: \newtheorem{lemma}[theorem]{Lemma}
18: \newtheorem{corollary}[theorem]{Corollary}
19: \newtheorem{proposition}[theorem]{Proposition}
20: \theoremstyle{definition}
21: \newtheorem{definition}[theorem]{Definition}
22: \newtheorem{example}[theorem]{Example}
23: \newtheorem{xca}[theorem]{Exercise}
24: \newtheorem{problem}[theorem]{Problem}
25: \theoremstyle{remark}
26: \newtheorem{remark}[theorem]{Remark}
27: \def\vgk{\text{\bf VG}_k}
28: \def\tN{\widetilde N}
29: \def\tL{\widetilde L}
30: \def\tM{\widetilde M}
31: \def\tpi{\widetilde \pi}
32: \def\hpi{\widehat \pi}
33: \def\hM{\widehat M}
34: \def\heatau{\lf(\frac{\p}{\p \tau}-\Delta\ri)}
35: \def\taub{\bar{\tau}}
36: \def\sigmab{\bar{\sigma}}
37: \def\texp{\widetilde \exp}
38: \def\heat{\lf(\frac{\p}{\p t}-\Delta\ri)}
39: \def \b {\beta}
40: \def\i{\sqrt{-1}}
41: \def\Ric{\text{Ric}}
42: \def\lf{\left}
43: \def\ri{\right}
44: \def\bbar{{\bar{\beta}}}
45: \def\a{\alpha}
46: \def\ol{\overline}
47: \def\g{\gamma}
48: \def\e{\epsilon}
49: \def\p{\partial}
50: \def\delbar{{\bar{\delta}}}
51: \def\ddbar{\partial\bar{\partial}}
52: \def\dbar{{\bar{\partial}}}
53: \def\tJ{\tilde J}
54: \def\C{\Bbb C}
55: \def\R{\Bbb R}
56: \def\P{\Bbb P}
57: \def\tx{\tilde x}
58: \def\vp{\varphi}
59: \def\tD{\tilde {\Delta}}
60: \def\k{\Upsilon}
61: \def\ty{\tilde y}
62: \def\H{\text{\rm Hess}}
63: \def\tE{\tilde{E}}
64: \def\tP{\tilde{P}}
65: \def\tI{\tilde{I}}
66: \def\dbar{\bar\partial}
67: \def\ba{{\bar{\alpha}}}
68: \def\bb{{\bar{\beta}}}
69: \def\abb{\alpha{\bar{\beta}}}
70: \def\gbd{\gamma{\bar{\delta}}}
71: \def\i{\sqrt {-1}}
72: \def\tD{\widetilde \Delta}
73: \def\Q{{\mathcal Q}}
74: \def\tn{\widetilde \nabla}
75: \def \D {\Delta}
76: \def\aint{\frac{\ \ }{\ \ }{\hskip -0.4cm}\int}
77: \def\id{\operatorname{id}}
78: \def\Ric{\operatorname{Ric}}
79: \def\I{\operatorname{I}}
80: \def\Scal{\operatorname{Scal}}
81: \def\tr{\operatorname{tr}}
82: \def\Ker{\operatorname{Ker}}
83: \def\Im{\operatorname{Im}}
84: \def\W{\operatorname{W}}
85: \def\l{\operatorname{l}}
86: \def\tri{\operatorname{tri}}
87: \def\tl{\tilde{\lambda}}
88:
89:
90: \numberwithin{equation}{section}
91: \begin{document}
92:
93: \title{Complete manifolds with nonnegative curvature operator }
94: \author{Lei Ni}\thanks{The first author was supported in part by NSF Grants and an Alfred P. Sloan
95: Fellowship, USA}
96:
97: %Information for first author
98:
99:
100: %Address of record for the research reported here
101:
102: \address{Department of Mathematics, University of California at San Diego, La Jolla, CA 92093}
103: % Current address
104:
105: \email{lni@math.ucsd.edu}
106:
107: %\thanks will become a 1st page footnote.
108:
109:
110: %Information for second author
111:
112:
113: %General info
114: \author{Baoqiang Wu}
115:
116: \address{Department of Mathematics, Xuzhou Normal University, Xuzhou, Jiangsu, China}
117:
118: \email{wubaoqiang@xznu.edu.cn}
119:
120: \date{June 2006}
121:
122: \begin{abstract}
123:
124: In this short note, as a simple application of the strong result
125: proved recently by B\"ohm and Wilking, we give a classification on
126: closed manifolds with $2$-nonnegative curvature operator. Moreover,
127: by the new invariant cone constructions of B\"ohm and Wilking, we
128: show that any complete Riemannian manifold (with dimension $\ge 3$)
129: whose curvature operator is bounded and satisfies the pinching
130: condition $R\ge \delta R_{I}>0$, for some $\delta>0$, must be
131: compact. This provides an intrinsic analogue of a result of Hamilton
132: on convex hypersurfaces.
133: \end{abstract}
134: \keywords{}
135:
136:
137:
138:
139:
140:
141:
142:
143:
144:
145:
146: \maketitle
147:
148: \section{Introduction}
149: Let $(M, g)$ be a Riemannian manifold. The curvature operator of $(M, g)$
150: lies in the subspace $S_B^2(\wedge^2TM)$ of
151: $S^2(\wedge^2TM)$ cut out by the Bianchi identity. The
152: decomposition $S_B^2(\wedge^2TM)=\langle \I\rangle \oplus \langle
153: \Ric_0\rangle \oplus \langle \W\rangle$ splits the space of algebraic curvature
154: operators into $O(n)$-invariant orthogonal irreducible subspaces. For an orthonormal
155: basis $\phi_{\alpha}$ (say $\phi_\alpha=e_i\wedge e_j$) of
156: $\wedge^2TM$ (which can be identified with $so(n)$), the Lie bracket
157: is given in terms of
158: $$
159: [\phi_\alpha, \phi_\beta]=c_{\alpha \beta \gamma} \phi_\gamma.
160: $$
161: It is easy to check, by simple linear algebra, that $ \langle [\phi,
162: \psi],\omega\rangle =-\langle [\omega, \psi], \phi\rangle. $ Here
163: $\langle A, B\rangle=-\frac{1}{2}\tr(AB)$. This immediately implies
164: that $c_{\alpha \beta\gamma}$ is anti-symmetric. If $A, B\in
165: S^2(\wedge^2TM)$ one can define
166: $$
167: (A\#B)_{\alpha \beta}=\frac{1}{2}c_{\alpha \gamma \eta}c_{\beta
168: \delta \theta}A_{\gamma \delta}B_{\eta \theta}.
169: $$
170: It is easy to see that $A\#B$ is symmetric too. Also from the
171: anti-symmetry of $c_{\alpha\beta\gamma}$
172: $
173: A\#B=B\#A.
174: $
175:
176: In \cite{BW}, a remarkable algebraic identity was proved on how a linear
177: transformation of $S_B^2(\wedge^2 TM)$ changes the quadratic
178: form $Q(R)=R^2+R^\#$.
179: B\"ohm and Wilking then constructed a continuous {\it pinching family of invariant closed convex
180: cones}. Using this construction they
181: confirmed a conjecture of Hamilton stating that {\it
182: on a compact manifold the normalized Ricci flow evolves a
183: Riemannian metric with $2$-positive curvature operator to a limit
184: metric with constant sectional curvature}. Hence it gives a
185: complete topological classification of compact manifolds with positive
186: $2$-positive curvature operator. In this short notes, based on the
187: strong result and the techniques of \cite{BW}, we give the
188: classification for manifolds with $2$-nonnegative curvature
189: operators and an application of their invariant cone constructions to
190: the compactness of Riemannian manifolds with pinching curvature
191: operator.
192:
193: \section{A strong maximum principle}
194: Let $(M, g(t))$ be a complete solution to Ricci flow such that
195: there exists a constant $A$ and the curvature tensor of $g(t)$
196: satisfies $|R_{ijkl}|^2(x,t) \le A$, for all $(x, t)\in M\times[0,
197: T]$. In \cite{H86}, Hamilton proved that under the evolving normal
198: frame the curvature tensor satisfies the following evolution
199: equation.
200: \begin{proposition}[Hamilton]
201: \begin{equation}
202: \label{ham861}\left( \frac{\partial}{\partial t}-\Delta\right)
203: R=2\left( R^{2}+R^{\#}\right)
204: \end{equation}
205: where $R^{\#}=R\#R$.
206: \end{proposition}
207:
208: The following was observed for compact manifolds in \cite{Chen,
209: H90}. We spell out the argument for the noncompact case for the sake
210: of the completeness.
211: \begin{proposition}\label{chen1}
212: The convex cone of $2$-nonnegative curvature operator is preserved
213: under the Ricci flow.
214: \end{proposition}
215: \begin{proof} Let $\I$ be the identity of $S_B^2(\wedge^2TM)$, which
216: can be identified with the induced metric on $\wedge^2TM$ (as a
217: section of $\wedge^2TM\otimes \wedge^2TM$). We also denote the
218: identity map of $TM$ by $\id$. With respect to the evolving normal
219: frame we have that $\nabla \I=0$ and $\frac{\p}{\p t}\I=0$. Let
220: $\psi(x,t)> 0$ be the fast growth function constructed in Lemma 1.1
221: of \cite{NT1} satisfying $\frac{\p}{\p t}\psi-\Delta \psi \ge C_1
222: \psi$. Here $C_1$ can be chosen as arbitrarily large as we wish. We
223: shall consider $\tilde R=R+\epsilon \psi \I$ and show that $\tilde R
224: $ is $2$-positive for every (sufficiently small) $\epsilon$. If not
225: by the boundedness of $R$ and growth of $\psi$ we know that it can
226: only fail somewhere finite. Assume that $t_0$ is the first time
227: $\tilde R$ fails to be $2$-positive and it happens at some point
228: $x_0$. If we choose orthonormal basis $\omega_\alpha$ (it may not be
229: in the form of $e_i\wedge e_j$ as $\phi_\alpha$) such that $\tilde
230: R$ is diagonal (so is $R$) with eigenvalue $\mu_1\le
231: \mu_2\le\cdot\cdot \cdot \le \mu_N$, where $N=\frac{n(n-1)}{2}$.
232: Parallel translate $\omega_\alpha$ to a neighborhood of $(x_0,
233: t_0)$, and let $\tilde R_{\alpha\alpha}=\langle R(\omega_\alpha),
234: \omega_\alpha\rangle$ then
235: at $(x_0, t_0)$ we have, by the maximum principle, that
236: \begin{eqnarray*}
237: 0&\ge& \left(\frac{\p}{\p t} -\Delta \right)\left(\tilde
238: R_{11}+\tilde R_{22}\right)\\
239: &\ge& (R^2+R^\#)_{11}+(R^2+R^\#)_{22}+2\epsilon C_1\psi\\
240: &=& (\tilde R^2+\tilde R^\#)_{11}+(\tilde R^2+\tilde
241: R^\#)_{22}+2\epsilon C_1\psi\\
242: &\quad &+\left(R^2+R^\#-\tilde R^2-\tilde
243: R^\#\right)_{11}+\left(R^2+R^\#-\tilde R^2-\tilde R^\#\right)_{22}
244: \\
245: &=&\mu_1^2 +\mu_2^2+\sum
246: (c^2_{1\beta\gamma}+c^2_{2\beta\gamma})\mu_\beta\mu_\gamma+2\epsilon
247: C_1\psi
248: \\&\quad &
249: -\epsilon \psi\left(\left(2\Ric\wedge \id +(n-1)\epsilon \psi
250: \I\right)_{11}+\left(2\Ric\wedge \id +(n-1)\epsilon \psi
251: \I\right)_{22}\right).
252: \end{eqnarray*}
253: Here in the last equation above we have used Lemma 2.1 of \cite{BW},
254: which asserts that $R+R\#\I=\Ric\wedge \id$ (the use is not really
255: necessary). Since $\mu_1+\mu_2\ge 0$ and $\mu_\gamma \ge 0$ for all
256: $\gamma\ge 2$,
257: $$
258: \sum
259: (c^2_{1\beta\gamma}+c^2_{2\beta\gamma})\mu_\beta\mu_\gamma=2\sum_{\gamma
260: \ge
261: 3}(c^2_{12\gamma}+c^2_{21\gamma})(\mu_1+\mu_2)\mu_\gamma+\sum_{\beta,
262: \gamma \ge
263: 3}(c^2_{1\beta\gamma}+c^2_{2\beta\gamma})\mu_\beta\mu_\gamma\ge 0.
264: $$
265: Notice also that at $(x_0, t_0)$ we have that $\mu_{11}+\mu_{22}=0$,
266: which implies that $R_{11}+R_{22}=-2\epsilon \psi$, then $2\epsilon
267: \psi(x_0, t_0)\le 2A$. Hence at $(x_0, t_0)$ we have that
268: \begin{eqnarray}\label{str1}
269: 0&\ge& \left(\frac{\p}{\p t} -\Delta \right)\left(\tilde
270: R_{11}+\tilde R_{22}\right)\nonumber\\
271: &\ge &\mu_1^2 +\mu_2^2+2\epsilon C_1\psi-200nA\epsilon \psi.
272: \end{eqnarray}
273: This is a contradiction if we choose $C_1>100nA$.
274: \end{proof}
275:
276: By choosing the barrier function more carefully as in \cite{NT2,
277: N04} (see for example Theorem 2.1 of \cite{N04}), we can have the
278: following strong maximum principle.
279:
280: \begin{corollary}\label{strong1} Assume that $R(g(0))$ is
281: $2$-nonnegative and $2$-positive somewhere. Then there exists
282: $f(x,t)>0$ for $t>0$ and $f(x,0)\ge \frac{1}{2}(\mu_1+\mu_2)$, such
283: that $$\left(\mu_{1}+\mu_{2}\right)(x,t)\ge f(x, t).$$ In
284: particular, if $R(g(0))$ is $2$-nonnegative and $(\mu_1+\mu_2)(x_0,
285: t_0)=0$ for some $t_0\ge 0$, then $(\mu_1+\mu_2)(x, t)\equiv 0$ for
286: all $(x, t)$ with $t\le t_0$. Moreover, $\mu_1(x, t)=\mu_2(x,t)=0$
287: for all $(x, t)$ with $t\le t_0$ and
288: $$
289: \mathcal{N}_2(x, t)=\mbox{span}\{\omega_1, \omega_2\}
290: $$
291: is a distribution on $M$ which is invariant under the parallel
292: translation.
293: \end{corollary}
294:
295: The above result together with (\ref{str1}) implies the following
296: classification of closed $2$-nonnegative manifolds.
297:
298:
299: \begin{corollary}\label{topo} Assume that $R(g(0))$ is
300: $2$-nonnegative. Then for $t>0$, either the curvature operator
301: $R(g(t))$ is $2$-positive, or $R(g(t))\ge 0$. Hence Suppose $\left(
302: M^{n},g_{0}\right) $ is a closed Riemannian manifold with
303: $2$-nonnegative curvature operator. Let $\tilde{g}\left( t\right) $
304: be the lift to the universal cover $\tilde{M}$ of the solution
305: $g\left( t\right) $ to the Ricci flow with $g\left( 0\right)
306: =g_{0}.$ Then for any $t>0$ we have either $\left(
307: \tilde{M}^{n},\tilde{g}\left( t\right) \right) $ is a closed
308: manifold with $2$-positive curvature operator or it is isometric to
309: the product of the following:
310:
311: \begin{enumerate}
312: \item Euclidean space,
313:
314: \item closed symmetric space,
315:
316: \item closed Riemannian manifold with positive curvature
317: operator,
318:
319: \item closed K\"{a}hler manifold with positive curvature operator on real
320: $\left( 1,1\right) $-forms.
321: \end{enumerate}
322: \end{corollary}
323: \begin{proof} It follows from the above corollary and Hamilton's classification result on the
324: solutions with nonnegative curvature operator. See for example
325: \cite{CLN}, Theorem 7.34.
326: \end{proof}
327:
328: Topologically, it is now known, by \cite{BW}, that simply-connected
329: $2$-positive manifolds is sphere, and the K\"ahler manifold in the
330: last case is biholomorphic to the complex projective space by the
331: earlier result of Mori -Siu-Yau. The fact that the curvature
332: operator of the evolving metrics becomes either $2$-positive or
333: nonnegative has been observed in \cite{Chen}. However, in
334: \cite{Chen} there is no clear statement of the strong maximum
335: principle, namely Corollary \ref{strong1}, on which the observation
336: relies. If evoking Theorem 2.3 of \cite{N04}, the splitting result
337: on solutions of Ricci flow on complete Riemannian manifold with
338: nonnegative curvature operator, we can write a similar statement
339: even when $M$ is not assume to be compact. However, in this case
340: the Euclidean factor is only topological (not isometric). Also we
341: do not know if a complete noncompact $2$-positive Riemannian
342: manifold is diffeomorphic to $\R^n$ or not.
343:
344:
345: \section{Manifolds with pinched curvature}
346:
347: In \cite{H91} Hamilton proved that any convex hypersurface (with
348: dimension $\ge 3$) in Euclidean space with second fundamental form
349: $h_{ij}\ge \delta \frac{\tr(h)}{n}\id$ must be compact. In
350: \cite{CZ}, using the pre-established estimates of \cite{Hu} and
351: \cite{Sh2}, Chen and Zhu
352: proved the following weak version of above-mentioned Hamilton's
353: result in terms of curvature operators. Namely, they
354: proved that {\it if a complete Riemannian manifold $(M^n, g)$ with
355: bounded and $(\epsilon, \delta_n)$- pinched curvature operator (with
356: $ n\ge 3$) in the sense that
357: $$
358: |R_{\W}|^2+|R_{\Ric_0}|^2\le \delta_n (1-\epsilon)^2|R_I|^2=\delta_n
359: (1-\epsilon)^2\frac{2}{n(n-1)}{\Scal (R)}^2
360: $$
361: for $\e>0$, $\delta_3>0, \delta_4=\frac{1}{5},
362: \delta_5=\frac{1}{10}$ and $\delta_n =\frac{2}{(n-2)(n+1)}$, where
363: $R_{\W}$, $R_{\Ric_0}$ and $R_{\I}$ denote the Weyl curvature
364: tensor, traceless Ricci part and the scalar curvature part. Then $M$
365: must be compact.} The strong pinching condition was the one
366: originally assumed in \cite{Hu} to obtain various estimates and the
367: smooth convergence result. It was also shown in \cite{Hu} that it
368: implies that $R\ge \epsilon R_{I}$. In \cite{N05} the first author
369: showed that the above result of Chen-Zhu can be shown by the blow-up
370: analysis of \cite{H90} and some non-existence results on gradient
371: steady and expanding solitons obtained in \cite{N05}. (The detailed
372: proof on these non-existence results were submitted to 2004 ICCM
373: proceedings a while ago. See also forthcoming book \cite{CLN}.) With
374: the help of a family of invariant cones constructed in \cite{BW}, we
375: can now prove the following general result.
376:
377: \begin{theorem}\label{ham}
378: Let $(M^n, g_0)$ be a complete Riemannian manifold with $n\ge 3$.
379: Assume that the curvature operator of $M$ is uniformly bounded
380: ($|R_{ijkl}|(x)\le A$) and satisfies that
381: \begin{equation}\label{pin1}
382: R\ge \delta R_{\I}>0
383: \end{equation}
384: for some $\delta>0$. Then $(M, g)$ must be compact.
385: \end{theorem}
386:
387: Recall that $R_{\I}=\frac{1}{n(n-1)}\Scal(R)\I$, where $\I$ is the
388: identity of $S_B^2(so(n))$. The above result is a natural analogue
389: of Hamilton result for hypersurfaces.
390: \begin{proof} Let $g(t)$ be the solution to Ricci flow with initial metric $g_0$ constructed
391: by \cite{Sh1}. First we show that if $M$ is noncompact, $g(t)$ can
392: be extended to a long-time solution defined on $M\times [0,
393: \infty)$. In order to do that we first show that for sufficient
394: small $b>0$, $R(g_0)$ lies inside the invariant cone constructed by
395: Lemma 3.4 of \cite{BW}. Recall from \cite {BW} the linear
396: transformation
397: $$
398: \l_{a, b}: R\to R+2(n-1)aR_{\I}+(n-2)b R_{\Ric_0}.
399: $$
400: More precisely
401: \begin{eqnarray*}
402: \l_{a, b}(R)&=&R+2a \bar{\lambda}\I +2b\id \wedge \Ric_0(R)\\
403: &=& (1+2(n-1)a)R_{\I}+(1+(n-2)b)R_{\Ric_0}+R_{\W}.
404: \end{eqnarray*}
405: It is easy to see that $\l_{a,b}(S_B^2(so(n)))\subset S_B^2(so(n))$
406: and is invertible if $a\ne -\frac{1}{2(n-1)}$ and $b\ne
407: -\frac{1}{n-2}$. Using this linear map and Theorem 2 of \cite{BW}, a
408: pinching family of invariant convex cones are constructed. In
409: particular, as one step of the construction, it was shown that
410: \begin{lemma}[B\"ohm-Wilking]\label{keylemma1}For $b\in [0, \frac{1}{2}]$, let
411: $$
412: a=\frac{(n-2)b^2+2b}{2+2(n-2)b^2}\, \mbox{ and }
413: p=\frac{(n-2)b^2}{1+(n-2)b^2}.
414: $$
415: Then the set $\l_{a, b}(C(b))$ where
416: $$
417: C(b)=\left\{R\in S_B^2(so(n))\, |\, R\ge 0,\, \Ric \ge
418: p(b)\frac{\tr(\Ric)}{n}\right\},
419: $$
420: is invariant under the vector fields $Q(R)$. In fact for $b\in(0,
421: \frac{1}{2}]$ it is transverse to the boundary of the set at all
422: boundary points $R\ne 0$.
423: \end{lemma}
424: We claim that there exists $b>0$ so small that $R(g_0)\in
425: \l_{a,b}(C(b))$, which is equivalent to that $\l_{a,
426: b}^{-1}(R(g_0))\in C(b)$. For simplicity let $\tilde R=R(g_0)$,
427: $\bar{\lambda}(\tilde R)=\frac{\Scal(\tilde R)}{n}$ and $\l=\l_{a,
428: b}$. Direct computation shows that
429: $$
430: R:=\l^{-1}(\tilde
431: R)=\tilde{R}_{\W}+\frac{1}{1+2(n-1)a}\tilde{R}_{\I}+\frac{1}{1+(n-2)b}\tilde{R}_{\Ric_0}
432: $$
433: which implies that
434: $$
435: \Ric(\l^{-1}(\tilde R))=\frac{\bar{\lambda}(\tilde
436: {R})}{1+2(n-1)a}\id+\frac{1}{1+(n-2)b}\Ric_0(\tilde R)
437: $$
438: and
439: $$
440: \bar{\lambda}(R):=\frac{\tr(\l^{-1}(\tilde
441: R))}{n}=\bar{\lambda}(\tilde R)\left(1-\frac{2(n-1)
442: a}{1+2(n-1)a}\right).
443: $$
444: Let $\tl_i$ be the eigenvalues of $\Ric_0(\tilde R)$. Then by the
445: assumption (\ref{pin1}) we have that
446: \begin{equation}\label{pin2}
447: \tl_i+\bar{\lambda}(\tilde R)\ge \delta \bar{\lambda}(\tilde R).
448: \end{equation}
449: Clearly we also have that
450: \begin{equation}\label{pin3}
451: \tl_i+\bar{\lambda}(\tilde R)\le n\bar{\lambda}(\tilde R).
452: \end{equation}
453: We first check that $R$ satisfies the Ricci pinching condition. In
454: fact if $\lambda_i$ are the eigenvalues of $\Ric_0(R)$, from the
455: above formulae we have that
456: \begin{eqnarray*}
457: -\lambda_i&=&-\frac{1}{1+(n-2)b}\tl_i\\
458: &\le& \frac{1-\delta}{1+(n-2)b}\bar{\lambda}(\tilde R)\\
459: &=& (1-\delta)\frac{1+2(n-1)a}{1+(n-2)b}\bar{\lambda}( R).
460: \end{eqnarray*}
461: Then there exist $\delta_1>0$ and $b_0$ such that for all $b\in [0,
462: b_0]$, $-\lambda_i\le (1-\delta_1)\bar{\lambda}( R)$. Then we can
463: find $b_1\le b_0$ such that for any $b\in [0, b_1]$, $p(b)\le
464: \delta_1$. Hence $R=\l_{a, b}^{-1}(\tilde R)$ satisfies the
465: pinching condition of $C(b)$. Now we check that $R=\l^{-1}_{a,
466: b}(\tilde R)\ge 0$. Rewrite
467: $$
468: R=\tilde{R}-\frac{2(n-1)a}{1+2(n-1)a}\tilde{R}_{\I}-\frac{(n-2)b}{1+(n-2)b}\tilde{R}_{\Ric_0}.
469: $$
470: Noticing that $a\to 0$ as $b\to 0$, we can find $b_2$ such that for
471: any $b\in [0, b_2]$ we have that
472: $$
473: R\ge \frac{\delta}{2}\tilde
474: R_{\I}-\frac{(n-2)b}{1+(n-2)b}\tilde{R}_{\Ric_0}.
475: $$
476: But the eigenvalue (with respect to $e_i\wedge e_j$, where
477: $\{e_i\}$ is a basis of $TM$ consisting of eigenvectors of
478: $\Ric_0(\tilde R)$) of the right hand side operator can be computed
479: as
480: $$
481: \frac{\delta}{2}\frac{\bar{\lambda}(\tilde
482: R)}{n-1}-\frac{b}{1+(n-2)b}\left(\tl_i+\tl_j\right).
483: $$
484: Using (\ref{pin3}), the above can be bounded from below by
485: $$
486: \bar{\lambda}(\tilde
487: R)\left(\frac{\delta}{2(n-1)}-\frac{2(n-1)b}{1+(n-2)b}\right)>0
488: $$
489: if $b$ is close to $0$. This shows that there exists $b_3>0$ such
490: that for any $b\in (0, b_3]$, $R(g_0)\in \l_{a, b}(C(b))$.
491:
492: Now the virtue of the proof of Theorem 5.1 in \cite{BW}, along
493: with the short time existence result of \cite{Sh1}, shows that the
494: Ricci flow has long time solution. Otherwise, by Theorem 16.2 of
495: \cite{H90}, we would end up with a blow-up solution, which is
496: nonflat, noncompact, but whose curvature operator $R=R_{\I}$. In
497: view of Schur's theorem, this is a contradiction. Note that
498: $R(g_0)\in \l_{a, b}(C(b))$ allows us to apply the generalized
499: pinching set construction (Theorem 4.1) from \cite{BW}, and since
500: the evolving metric has positive curvature operator and the manifold
501: is assumed to be noncompact, the injectivity radius always has a
502: lower bound in terms of the size of the curvature. All these
503: ingredients allow us to perform Hamilton's blow-up analysis
504: \cite{H90} (Theorem 16.2).
505:
506: We continue to show that the extra assumption that $M$ is
507: noncompact
508: will lead us to a contradiction by performing the singularity analysis of
509: \cite{H90} as $t\to \infty$. Notice that for all $t$, $R(g(t))$
510: will stay in the cone $\l_{a, b}(C(b))$ for some fixed (but
511: sufficiently small) $b$, by the tensor maximum principle, which can
512: be verified in the same way as Proposition \ref{chen1}. Now we claim
513: that
514: the curvature of $g(t)$ satisfies that
515: \begin{equation}\label{ric-pin}
516: \Ric \ge p\frac{\tr(\Ric)}{n}\id.
517: \end{equation}
518: for some $p>0$. Let $R^*=R(g(t))$. First, by Lemma \ref{keylemma1}
519: we know that $R(g(t))\in \l_{a,b}(C(b)$ for some fixed small $b$.
520: Thus we can find $R\in C(b)$ such that $\l_{a, b}(R)=R^*$. Now let
521: $\bar{\lambda}=\frac{\tr(\Ric(R))}{n}$ and $\lambda_i$ be the
522: eigenvalues of $\Ric_0(R)$. By the assumption we have that
523: $-\lambda_i\le (1-p)\bar{\lambda}$. Now we compute the Ricci
524: curvature and its trace for $R^*$. By the definition of $\l_{a, b}$
525: we have that
526: $$
527: \Ric(R^*)=\Ric +2(n-1)a \bar{\lambda}\id +(n-2)b \Ric_0
528: $$
529: and
530: $$
531: \bar{\lambda}^*:=\frac{\tr(\Ric(R^*))}{n}=\bar{\lambda}(1+2(n-1)a).
532: $$
533: Letting $\lambda^*_i$ be the eigenvalue of $R^*$ we have that $
534: \bar{\lambda}^*+\lambda_i^*
535: =(1+2(n-1)a)\bar{\lambda}+(1+(n-2)b)\lambda_i $. Therefore
536: \begin{eqnarray*}
537: -\lambda_i^*&=&-(1+(n-2)b)\lambda_i\\
538: &\le &(1-p)(1+(n-2)b)\bar{\lambda}\\
539: &=& (1-p)\frac{1+(n-2)b}{1+2(n-1)a}\bar{\lambda}^*\\
540: &\le &(1-p)\bar{\lambda}^*.
541: \end{eqnarray*}
542: Here we have used the fact that
543: $1+2(n-1)a=1+(n-1)\frac{(n-2)b^2+2b}{1+(n-2)b}>1+(n-2)b$. This
544: completes the proof of the claim (\ref{ric-pin}).
545:
546:
547: Since for all $g(t)$, its Ricci curvature satisfies (\ref{ric-pin}),
548: this holds up on the blow-down/blow-up solutions, which after
549: passing to its universal cover, are either a nonflat gradient steady
550: soliton or a nonflat gradient expanding soliton, with nonnegative
551: curvature operator, by results from \cite{H90} (Theorem 16.2,
552: Corollary 16.4) (See also \cite{N02}, Theorem 4.2 and \cite{CZ}).
553: This contradicts to Corollary 3.1 of \cite{N05}.
554: \end{proof}
555:
556:
557: \section{Discussions}
558:
559: In \cite{W}, the topology of so-called $p$-positive manifolds was
560: studied. In view of the result of B\"ohm-Wilking, it is reasonable
561: speculate that any noncompact complete Riemannian manifold with
562: $2$-positive curvature operator must be diffeomorphic to $\R^n$. In
563: \cite{N05} we speculated that any complete Riemannian manifolds with
564: positive pinched Ricci curvature must be compact. Theorem \ref{ham}
565: confirms it under stronger assumption on curvature operator. The
566: problem in full generality still remains unknown.
567:
568:
569: \medskip
570:
571: {\it Acknowledgement}. Part of this paper was completed during the
572: first author's visit of ETH, Z\"urich. He would like to thank ETH,
573: especially Tom Ilamnen, for providing a stimulating environment and
574: various discussions. He also held informal discussions on \cite{BW}
575: with Ben Chow and Nolan Wallach.
576:
577:
578: \begin{thebibliography}{9999}
579:
580: \bibitem[BW]{BW} C. B\"ohm and B. Wilking, \textit{ Manifolds with positive curvature operator are space form},
581: preprint.
582:
583: \bibitem[Chen]{Chen} H. Chen, \textit{ Pointwise $\frac14$-pinched $4$-manifolds}, Ann. Global
584: Anal. Geom. 9 (1991), no. 2, 161--176.
585:
586: \bibitem[CZ]{CZ} B.-L. Chen and X.-P. Zhu, \textit{ Complete
587: Riemannian manifolds with pointwise pinched curvature} Invent. Math.
588: \textbf{ 140}(2000) no. 2, 423--452.
589:
590: \bibitem[CLN]{CLN}B. Chow, P. Lu and L. Ni, \textit{ Hamilton's Ricci flow,}
591: Graduate Studies in Mathematics, AMS Press, to appear.
592:
593:
594: \bibitem[H1]{H86} R. Hamilton, \textit{ Four-manifolds with positive curvature operator }, J. Differenital.
595: Geom. \textbf{24} (1986 ), 153--179.
596:
597: \bibitem[H2]{H91} R. S. Hamilton,\textit{
598: Convex hypersurfaces with pinched second fundamental form,} Comm.
599: Anal. Geom. \textbf{2}(1994), no. 1, 167--172.
600:
601: \bibitem[H3]{H90} R. S. Hamilton,\textit{ Formation
602: of singularities in the Ricci flow,} Surveys in Differential
603: Geom.\textbf{ 2}(1995), 7--136.
604:
605:
606: \bibitem[Hu]{Hu}G. Huisken, \textit{ Ricci deformation of the metric on a Riemannian
607: manifold,} J. Differential Geom. \textbf{21}(1985), no. 1,
608: 47--62.
609:
610:
611:
612: \bibitem[N1]{N04} L. Ni, \textit{ Ricci flow and nonnegativity of sectional curvature}, Math. Res. Lett.
613: \textbf{11}(2004), no. 5-6, 883--904. %
614:
615: \bibitem[N2]{N02}L. Ni, \textit{ Monotonicity and K\"ahler-Ricci flow. Geometric evolution
616: equations,} 149--165, Contemp. Math., \textbf{367}, Amer. Math.
617: Soc., Providence, RI, 2005.
618:
619: \bibitem[N3]{N05} L. Ni, \textit{Ancient solutions to K\"ahler-Ricci flow}, Math. Res. Lett. \textbf{12}(2005),
620: no. 5-6, 633--653.
621:
622: \bibitem[NT1]{NT1} L. Ni and L.-F. Tam, \textit{Plurisubharmonic functions and the
623: K\"ahler-Ricci flow,} Amer. J. Math. \textbf{125}(2003), no. 3,
624: 623--654.
625:
626: \bibitem[NT2]{NT2}L. Ni and L.-F. Tam, \textit{ Plurisubharmonic functions and the structure
627: of complete K\"ahler manifolds with nonnegative curvature}, J.
628: Differential Geom. \textbf{64}(2003), no. 3, 457--524.
629:
630: \bibitem[Sh1]{Sh1} W. X. Shi, \textit{ Deforming the metric on complete
631: Riemannian manifolds,} J. Differential Geom.\textbf{30}(1989),
632: 223--301.
633:
634:
635: \bibitem[Sh2]{Sh2} W. X. Shi,\textit{ Ricci deformation of the metric on complete
636: noncompact Riemannian manifolds,} J. Differential
637: Geom.\textbf{30}(1989), 303--394.
638:
639:
640: \bibitem[W]{W}
641: H. Wu, \textit{Manifolds of partially positive curvature,} Indiana
642: Univ. Math. J. \text{36}(1987), no. 3, 525--548.
643:
644: \end{thebibliography}
645:
646: \end{document}
647: